Prannoy Roy has been synonymous with elections since 1980. He pioneered opinion polls in India and introduced psephology to the country. His polls and commentaries are valued for their objectivity and precision. He is also the co-founder of NDTV.
Dorab R. Sopariwala, an editorial adviser with NDTV, is a fellow of the Royal Statistical Society and has been involved with opinion polling for four decades now.
Roy and Sopariwala have made TV viewing an exciting experience during election season with their reporting, number-crunching and analysis. They have distilled their experience in decoding the Indian elections in a new book, The Verdict: Decoding India’s Elections . Excerpts from an interview:
You make this point in the book about the hidden conservative voter. Given all the noise that the Indian right wing makes, does it hold true for our country too?
PR: Probably not. From my experience, in fact, it is the opposite. In the sense that the main support of the BJP comes from the middle and upper-middle sections of the society, and from men. They are generally much more vocal. They will always turn up in the crowd and quite volubly share their views and that kind of intimidates people around. Among the poorest, the Scheduled Castes, and especially the Muslims, there is a fear factor. Once we went into a Dalit village, and we were generally chatting with people there. We asked them who they were voting for, and they said BJP, BJP… We spent a lot of time with them, had tea, and after a while, they all started laughing. We asked them why. ‘What we told you earlier was all nonsense… We thought you were some BJP politicians or something… But none of us is BJP.’ In India, there is no hidden conservative voter.
Is there a risk in India of over-reporting BJP support?
PR: Yes. We have found that in our surveys.
DS: In America for instance, the upper-class conservative voters are the ones who are quiet. The lower class, factory workers and all, are very vocal. In India, the upper-class conservative voters are not all quiet while the lower caste voters in general are muted.
Alliances are very critical in this election. When parties make an alliance, do all their respective voters add up to form one unit, or is there anything beyond arithmetic at play?
PR: When parties form alliances, not only do their voters come and vote for the alliance, they in fact get an additional share of votes because of the momentum they have created by forming the alliance.
DS: So, when the PMK contests all alone in Tamil Nadu, it might not win any seats. But when it is in an alliance — we are not talking about the Communists and the BJP fighting together, we are talking of parties that have virtually no ideological differences but are dominated by different leaders — they get an extra bump, as people perceive they have a higher chance of winning. So two plus two could become five in the case of alliances.
PR: There is this conventional wisdom that Mayawati’s supporters will vote for SP but SP’s supporters will not vote for BSP. But there is no evidence for that at all.
You mention in the book how most elections in India have been landslides. Do you expect 2019 to be a landslide too?
PR: When we talk of landslides we are talking of State-level results. The results of the Lok Sabha as a whole may not be a landslide, but in each State, it could be a one-sided affair 77% of the time. So, whenever you are doing that exercise, whoever you think is the winner in each State, the chances of that party winning a landslide is much higher than winning narrowly. Landslides are more common in Lok Sabha elections State-wise, but not overall landslides. For example, in TN, one side sweeps and then the other side sweeps.
You talk of ‘homogeneous swing zones’ — regions that show similar voting behaviour. Do the regions that voted the BJP in 2014 — western, northern and central India — constitute a homogeneous swing zone?
PR: Homogenous swing zones are smaller than States — within States, we often have two or three homogenous swing zones. Let’s take Karnataka. Southern Karnataka is Vokkaligas and north is Lingayats. They tend to go differently in terms of their swings. We identify homogeneous swing zones not based on geography of caste, but on data. Once we see a swing zone, we see the underlying social or economic factors at work there.
Though the BJP appears to be doing uniformly well across multiple States, there are regions within these States that have distinct political behaviour...
PR: Exactly.
DS: For example, eastern UP and western UP have very different social compositions. What motivates eastern UP may not touch western UP. In Andhra Pradesh, Rayalaseema, Telengana and coastal Andhra used to be different swing zones.
How sound is the notion of a rural-urban divide as a national phenomenon in understanding voter behaviour?
PR: Again, there is nothing like a national phenomenon. The Lok Sabha election is a federation of State elections. Each State is very different from another. In some States there is an urban-rural divide, in some others it is not there. Overall, the only thing that you can say from the data is that the rural turnout is much higher than urban. Rural women are voting in much higher numbers than urban women. There is no pattern across States of party-wise differentiation between urban and rural voters.
So what was called a Modi wave in 2014 was an aggregate of several regions where the BJP did well?
PR: When you talk about the Modi wave, I think it is a misnomer. He got a lot of seats, but the party won 31% of the votes. And the predominant reason for that victory was a divided opposition. Nowhere in the world will it be called a wave.
